Been working on a new driver for higher powered LEDs. It's a buck converter and has a nominal 8 - 24V input range.
Output is up to 16V and so far I've tested it at up to 2A output driving a 6V load (a Lamina Red BL2000) for a few hours.
I'll be cranking it up a bit beyond that soon. It is fully current regulated and dimmable via an onboard trimpot or external pot (similar to Fatman).
Here's a pic of a prototype next to a Star for size comparison.
And a pic of another proto driving a BL2000 at 2A. The BL2000 IS bright - the camera shuttering down to 1/200 sec ISO 100 [img]/ubbthreads/images/graemlins/grin.gif[/img]
[ QUOTE ] jtice said:
muuuuhahahahahaha more power [img]/ubbthreads/images/graemlins/smile.gif[/img]
another great looking driver george! [img]/ubbthreads/images/graemlins/thumbsup.gif[/img]
Are you still planning on making the driver for the BL3000 ?
hey. BTW: hows the Fat-Flax doing? I have mag mods to make man [img]/ubbthreads/images/graemlins/wink.gif[/img]
[/ QUOTE ]
Thanks, on the next protorun I'm going to stick an nFlex core on the back that could be 'optionally' populated. There's nothing (but copper) on the back of the pcb right now.
Yeah, still planning on production BL3000 drivers. I've made several protos and they work great. It's just waiting on some financial committment before I run a batch of boards & order up some qty of components.
I have the proto of the fat-flex running - needs some 'tweaking' so I'll most likely do a new pcb layout in the next week or two to test some changes.
Too many projects, too little time [img]/ubbthreads/images/graemlins/grin.gif[/img]
sign me up.. you can count on me picking up at least one of whatever you are coming up with george.. and with any luck.. 1000 or so of that special model i emailed about :-D.
Hehehe, I might be talking with ya about a whole mess of Fatmen for a project I'm working up, but in the meantime color me interested. [img]/ubbthreads/images/graemlins/grin.gif[/img]
Well, been a while, but I ran some new prototype pcbs through and got them back today. So, I put a few drivers through with a new layout. Same basic specs, buck converter, runs from up to 24V input, 85% to 94% efficiency depending on input/output voltage differential and output current. From 12V input, driving 2A into the Lamina array (~6.5V output) the efficiency is ~90%. Capable of up to 2.8A output. Fully current regulated, with on board 25 turn trimpot to adjust the output current. The board is 1.7" x 1.1" with components only on the top side.
Here it is running about 50mA into a Lamina Ceramics BL2000 red. The BL2000 Red has a typical Vf of 6.7V and Drive current of 2.8A.
A few turns of the trimpot and it's pushing about 2A into the same array.
Anybody ever use a CC driver for operating incan lights... 'hotwire'.. i just started dabbling with hotwire and they are fun but dim too quickly for my taste.. i would put a zener with a resistor in parallel with the output of the driver for open-load protection when the bulb burns out.. i think that with constant current it would be nice to limit the ludicrous inrush of current on a cold bulb.. the main problem i was having with trying a prototype of this was the relatively high currents required compared to LEDs.. but this driver looks really promising.. you have a high-power boost driver coming also, correct? I need that one for my 7 emitter upgrade of my megasonic i want to build soon.
[ QUOTE ] chimo said:
George, that looks great! A quick question: will the driver board require any heatsinking of its own at max output?
Paul
[/ QUOTE ]
I've run the BL2000 (~7V Vf at 2.5A from 14V input) for extended periods and the driver has remained just warm sitting on my bench. The back of the board is mostly all ground plane and the switcher is thermally coupled to it, so there is quite a bit of area for heat to spread/radiate. Worst case the back of the board could be thermally epoxied to a heatsink/body.
To determine efficiency (hence power dissipation) - one needs to consider, input voltage, output voltage, output current.
Really looking forward to getting a BL2000 and BL3000 soon, combined with one of your future boards,
should make a mean pelican case light [img]/ubbthreads/images/graemlins/grin.gif[/img]
I like the 25 turn pot, mush better than the flaky fatman pots,
though they are fine for a one time adjustment.
Are you still planning on adding the nflex core to them?
Jtice,
well, the "nice" 25 turn pot would be a little large for Fatman [img]/ubbthreads/images/graemlins/grin.gif[/img]
I could add the nFlex core to this driver, but I'll bring out the "dumb" version first since for many Lamina folk, they just want a fixed output driver.
And, for folk that want to know where to get the Lamina arrays - you can buy them online from Mouser or Digikey.
Thanks, George. That's great news. I noiced that you added a bit more capacitance from the prototype. This driver looks ideal for automotive apps and some large SLA flood/spot lamps (as well as other). [img]/ubbthreads/images/graemlins/clap.gif[/img]
[ QUOTE ] chimo said:
Thanks, George. That's great news. I noiced that you added a bit more capacitance from the prototype. This driver looks ideal for automotive apps and some large SLA flood/spot lamps (as well as other). [img]/ubbthreads/images/graemlins/clap.gif[/img]
Paul
[/ QUOTE ]
Yes, I added the "correct" amount on the new proto (good observation). Yes, this would be a great driver in an auto application - one of its intended functions [img]/ubbthreads/images/graemlins/grin.gif[/img]
I never answered the question about ~$ - I'm guessing around the $22 mark, depends on the volume that I order boards/parts in.
[ QUOTE ] andrewwynn said:
Anybody ever use a CC driver for operating incan lights... 'hotwire'..-awr
[/ QUOTE ]
Hi Andrew - I cannot remember all of the reasons, but constant current regulators will make an incan filament die fast. (according the experts in that area). I think it has to do with the highly non linear behavior of filament resistance. I think you need to look for constant voltage regulators.
i am thinking of making a constant voltage supply.. i could always convert one of the CC drivers it should not be difficult.. vreg is much easier than c.reg. The reason i thought CC would be good is that the startup current would not spike like typical .. but the voltage would swing around.. i think i'll be doing a test of CC and see what happens.. but i'd be very interested in getting in-touch with somebody that already did this testing.
a pair of 168s doesn't have enough energy but it does sound neat.. nice light for the short runtime it'd last. the 9AAs i have is about twice the WH of two 168s.. so i guess i might be building a driver or converting the likes of this one to voltage v. current.
New buck driver prototype
Reply With Quote
